Multi-agent aerial robotics systems are widely studied topics that present several challenges, including nonlinear dynamics, limited onboard resources, communication constraints, and environmental uncertainties. Due to these challenges, the design of control and navigation systems requires a careful integration of modeling, planning, and control strategies to ensure reliable and stable operation at the single-agent level, as well as effective coordination in multi-agent scenarios. The aim of this course is to introduce the students to the design of control and planning algorithms for autonomous systems, with a focus on unmanned aerial vehicles and their distributed coordination. The course covers fundamental and advanced topics in modeling, motion planning, and control of quadrotor UAVs, and progressively extends these concepts to multi-agent systems, including consensus algorithms and networked control. Particular attention is given to multi-UAV coordination problems, which represent a rapidly evolving research area and a source of ongoing developments in the field of autonomous systems.

Basic knowledge of control theory and autonomous control is beneficial for understanding the course's core concepts.

The first part of the lectures will then introduce the main concepts of aerial robotics and autonomous navigation, together with the mathematical modeling of quadrotor UAVs, highlighting the key features of their nonlinear dynamics and control challenges. The course will address motion planning strategies, including graph-based and sampling-based approaches such as RRT* and its variants, as well as optimization-based methods like MPPI, and will discuss trajectory generation and tracking for single-agent systems. Practical aspects related to system development will also be considered, including the use of ROS2 and simulation tools for control design and validation. Building on these foundations, the course will explore the distributed coordination of autonomous systems, with a specific focus on consensus control. Through concepts borrowed from graph theory, it will be explained how to perform distributed tasks in first- and second-order linear systems, both in continuous and discrete time. Advanced topics will include the analysis of networks with communication delays, switching topologies, and the design of distributed observers and filters for networked systems. Finally, distributed coordination strategies will be addressed in the context of multi-UAV systems, considering cooperative tasks under realistic constraints such as communication delays and limited information exchange.

Modalità mista

Presentazione orale

P.D.2-2 - Maggio